Background & Aims There is an unmet need for immunological biomarkers in chronic hepatitis B (CHB), where patient management relies on virological and biochemical markers despite the crucial role of virus-specific T cells in controlling viral replication and disease progression. To address this, we developed the HBV-cytokine release assay (HBV-CRA), a rapid, point-of-care test that measures multiple cytokines in whole blood after HBV-peptide stimulation.
Methods We first assessed the assay’s sensitivity by spiking whole blood with engineered HBV-specific T cells. Next, we compared sensitivity and consistency of HBV-CRA to ex vivo IFN-γ ELISpot assays. We then applied the assay in a cross-sectional study of 235 CHB patients and longitudinally in acute HBV patients during HBsAg sero-clearance.
Results The HBV-CRA detected T cell function in 80% of CHB cases and showed that elevated IL-2 and IFN-γ levels after Core peptide stimulation were associated with HBsAg clearance. Importantly, unsupervised clustering identified distinct immune response patterns independent of established clinical and virological classifications. The assay also demonstrated the functional impact of NUC treatment on HBV-specific T cell responses.
Conclusions The HBV-CRA is a rapid and easy-to-use assay that identifies immune profiles associated with HBsAg clearance and differentiates CHB patients based on antiviral T cell function. Its application in a large CHB cohort revealed that traditional disease phase classifications, based on viral and clinical parameters, cannot predict HBV-specific T cell profiles. The HBV-CRA has the potential to guide patient stratification for immunotherapeutic interventions.
Functional and quantitative defects of HBV-specific T cells are likely the primary reason chronic HBV infection (CHB) patients cannot achieve HBsAg seroconversion. Research showed that a high frequency of HBV-specific CD8+ T cells in the bloodstream, observed only in acute but not chronic HBV infection, leads to HBsAg loss and a 90% reduction in HBV-DNA levels. Furthermore, while acute HBV patients possess functional HBV-specific Th1-like CD4 T cells, CHB patients exhibit dysfunctional HBV-specific CD4 T cells, which likely contribute to the failure of HBs-specific B cells to fully mature and produce anti-HBs antibodies.
While acute HBV patients show a consistent profile of robust antiviral immunity, the adaptive immune defects in CHB patients are highly variable. HBV-specific T cells in CHB are present at low frequencies (<0.1% of T cells) and show heterogeneity in exhaustion marker expression, including PD-1, TIM-3, Lag-3, and CD160. Functionally, these T cells exhibit impaired proliferation and cytokine production. Their defects are heterogeneous, not exclusively associated with the expression level of inhibitory receptors and not solely related to classical T cell exhaustion mechanisms due to persistent antigenic stimulation. These defects vary with HBV antigen specificity, interaction with liver endothelial cells and their “priming history”, as T cells primed by hepatocytes or dendritic cells possess distinct functional profiles. Although better HBV-specific T cell function can sometimes be observed in CHB patients with low viral load or younger age, we cannot interpret the intricate spectrum of HBV-specific T cells through the standard virological and clinical assessments like HBsAg, HBeAg status or ALT levels. Only advanced immunological methods can define the complex nature of the HBV-specific T cells in CHB patients. Still, these methods are limited to highly specialized laboratories and not easily translated into clinical practice for large clinical trials.
Thus, we developed a simple, point-of-care immunological test that can define the functional profile of HBV-specific T cells of acute HBV patients during the dynamic phase of HBV control and the complex heterogeneity of HBV-specific T cells in CHB patients.
We modified a method that we developed to rapidly measure the frequency and function of SARS-CoV-2-specific T cells. We designed HBV peptide pools covering multiple genotypes and used them directly in the blood of acute HBV and CHB patients to stimulate the release of cytokines from HBV-specific T cells. After overnight incubation, IFN-γ, IL-2, IL-10, granzyme B (GrzB), TNF-α, and IL-5 were measured. We demonstrated that this simple HBV-cytokine release assay (HBV-CRA) enables the quantitative and functional profile of HBV-specific T cells using minimal blood volume. By applying the HBV-CRA in a large cohort of 235 CHB patients, we identified distinct HBV-specific T cell functional profiles, or secretomes, even among patients classified within the same clinical and virological categories, allowing their immune-based stratification.
Clinical assessment and blood sampling were performed at The Royal London Hospital, Barts Health NHS Trust, Singapore General Hospital and Policlinico of Milan in accordance with the Declaration of Helsinki. Written informed consent was obtained, and the Institutional Review Boards approved the study (refs: 17/LO/0266, 2022/2732, ID-4670).
260 individuals were included in this study: HBV-vaccinated subjects (n=16), treatment-naïve CHB patients (n=100), NUC-suppressed CHB patients (n=74), HBsAg-negative CHB patients (n=61), and patients with acute resolving HBV infection (n=9). A subset of NUC-suppressed patients (n=10) was sampled monthly over three time points. Acute HBV patients were tested 1-4 times during viral clearance. Tables S1, S2 and S3 summarized the clinical and virological parameters.
Table S1:Clinical and virological characteristics of healthy subjects and NUC-suppressed CHB patients followed monthly
Table S2:Clinical and virological characteristics of acute HBV patients in this study
HBV-CRA was performed on indicated samples. The sum of IFN-γ and IL-2 secreted in response to the six HBV-peptide pools (Core, S, PreS, Pol-1, Pol-2, X) is shown.
Table S3:Clinical and virological characteristics of all CHB patients in this study
We used a library of 838 overlapping 15-mer peptides (10-aa overlap). These peptides were combined into six pan-genotype pools covering all HBV proteins from genotypes A, B, C, and D and sourced from T Cell Diagnostics. SARS-CoV-2-Spike peptide pool was generated as described before.
PBMCs were seeded at 400,000 cells per well onto anti-IFN-γ (1-D1K, Mabtech; 5 μg/mL) pre-coated 96-well plates. Cells were stimulated for 18h with peptide pools (2 μg/mL), DMSO, or PHA controls. Plates were developed using biotinylated IFN-γ detection antibody (7-B6-1, Mabtech; 1:2000), Streptavidin-AP (Mabtech), and KPL BCIP/NBT substrate (SeraCare). We quantified Spot-forming cells (SFC) with ImmunoSpot, and positive responses were calculated by subtracting 2x SFC of DMSO controls from peptide-stimulated wells. All controls met quality criteria, with <5 SFC in negative controls and positive PHA controls.
HLA-A2-restricted HBV-Env 183-91-specific T cells were generated by electroporating mRNA encoding the TCR Vα and Vβ chains into 10-day expanded T cells from an HLA-A2+ subject. HBV-Env 183-91-TCR expressing T cells were quantified via dextramer staining.
On the same day of collection, whole blood drawn into heparinized tubes (320 μL blood per peptide pool) was mixed with 80 μL RPMI and stimulated with HBV peptide pools (2 μg/mL) or DMSO control. After 16 hours of culture at 37°C, the culture supernatant (plasma) was collected and stored at -80°C for cytokine quantification.
We measured IFN-γ, IL-2, TNF-α, IL-4, IL-5, IL-10, and GrzB in the plasma using an Ella machine (ProteinSimple). Cytokine levels in DMSO controls were 2x subtracted from peptide-stimulated samples. Dimensionality reduction was performed using UMAP and Phenograph with nearest neighbors = 200, min_dist = 0.99, k = 30. UMAP results were converted to .fcs files and analyzed in FlowJo to generate cytokine secretion heatmaps.
All tests are stated in the figure legends. P-values (all two-tailed): <0.05 = *; <0.01 = **; <0.001 = ***; <0.0001 = ****.
During the COVID-19 pandemic, we demonstrated that whole blood stimulation with peptide pools covering SARS-CoV-2 proteins can detect SARS-CoV-2-specific T cells more accurately than other assays (ELISpot, AIM assay). However, while vaccine- or infection-induced SARS-CoV-2-specific T cells often comprise around 0.1% of total T cells, circulating HBV-specific T cells in CHB patients rarely reach this level. This different magnitude is depicted in Figure 1A, where we tested T cells of COVID-19-vaccinated CHB patients against SARS-CoV-2 Spike and HBV-proteins. While the ex vivo IFN-γ ELISpot assay detected SARS-CoV-2-specific T cells, HBV-specific T cells were rarely found. To assess whether the whole blood HBV-CRA can achieve a sensitivity threshold suitable for detecting the low frequency of HBV-specific T cells, we diluted graded numbers of T cells with known specificity into whole blood. We engineered a T cell receptor (TCR) specific for the HLA-A2-restricted HBV-Env 183-91 epitope into T cells of an HLA-A2+ subject and we diluted decreasing numbers of HBV-Env 183-91-TCR expressing T cells into the whole blood from the same subject (Figure 1B). After overnight culture with the corresponding HBV-Env-183-91 peptide, plasma cytokines were measured. Figure 1C shows that IFN-γ and GrzB levels >10 pg/mL were detected when 50-100 Env 183-91-TCR+ T cells (∼0.01-0.02% of total T cells) were diluted in blood, matching the frequency of circulating HBV-specific T cells in most CHB patients. IL-2 was first detected at ∼100 cells/320 μL of blood, but levels >10 pg/mL were observed only at higher T cell numbers (∼1000), in line with the effector profile of these engineered TCR-T cells.
